Controlling curiosity


I find myself treating the experience like a video game: starting out cautiously with a few gentle turns, befor...


I find myself treating the experience like a video game: starting out cautiously with a few gentle turns, before commanding the car-sized rover to cross some sizable boulders. Mobility test engineer Daniel Fuller calls _game over_ when I manage to ground one of the vehicle_s six wheels on a particularly large lump of stone.

Thankfully the real rover, some 200 million kilometers (125 million miles) away on the Red Planet, is in safer hands. And the way it_s driven is rather different.

Related: Rover feeling good

_It_s a common misconception that we_re just _joysticking_ it or we_re driving and sending these commands and we_re suffering a 15 minute delay,_ says Fuller, who divides his time between the rover control room and Mars yard. _We actually plan an entire day _ or a sol as it_s called on Mars._

Each of these daily missions is overseen by people like Sanjeev Gupta from Imperial College London. When I meet him, the geologist and senior member of the JPL science team is living on Mars time.

_During the Mars night we prepare the sequences the rover has to do,_ he explains. _Then we uplink those commands and, when the rover wakes up, it carries out those tasks. But a Mars day is 40 minutes longer than an Earth day so you get progressively out of synch, so the time I arrive at work changes by an hour every day._

And that_s not all. _We don_t get a signal direct from the rover to Earth,_ says Gupta. _It has to go through communications satellites - Mars Reconnaissance Orbiter (MRO) and Odyssey - and these obviously have different times [when they_re overhead]._

_Generally you_re moving an hour forward each day but suddenly, like tomorrow, I_m going to step back two hours. I_m progressively getting used to it but the first three weeks were awful,_ he groans, _I was permanently exhausted._

Unexpected plans

During the Martian night, dozens of scientists, including Gupta, anaylse the images and data coming back from the Red Planet and discuss what they want to do next. They hold a series of meetings to prioritise time on the instruments, choose which images to capture and where the rover_s going to move.

_It_s quite complicated because you have to work out how much power each instrument is going to use and how much heating time they require,_ Gupta says. _Because they_re so many instruments, everybody wants to have a go _ so one instrument might get five minutes, another ten minutes._

Once the arguments are had and decisions reached, the drivers get to work, coding the next day_s instructions. And this is way more complicated than programming one of those _80s-era Big Trak robot toys.

Curiosity is the fourth rover (since the 1996 Pathfinder mission) Nasa has run on Mars in recent years, and the procedures have got better with every mission. Armed with 3D images from its navigation cameras, the drivers work through the Martian night to map out the best, smoothest, route to the next destination. They factor in any stops along the way, to take pictures or operate an instrument, and run a simulation of the journey to double-check they have got it right.

The aim is that when a new day dawns on Mars, the drivers are ready to go. Once either the Odyssey or MRO satellite is in the right place, they send a command to wake up the rover and upload its commands for the day. A bit like emailing a _to do_ list.

Gupta admits that the whole process is more time consuming than he first imagined. _Things that I can do in an hour on Earth, take days on Mars. Usually on Earth I work by myself, here we_ve got a team of 200 scientists so we have to work out these interactions and protocols._

Once it_s started moving, Curiosity also has the ability to think for itself using data from its hazard avoidance cameras and built-in safely protocols. This autonomy should prevent it grounding on a rock or falling over an unexpected cliff but also adds to the pressure on the drivers and scientists planning each sol_s activities.

This is where the rovers in Mars Yard come in handy. This dusty outdoor arena at the back of the lab is around the size of a couple of tennis courts. Strewn with boulders, mounds of sand and even a cliff face, it_s home to two full-scale mock-ups of Curiosity. Although the real thing was tested extensively before launch, these rovers are still in use everyday by drivers trying to choreograph movement on Mars.

_We do have a pretty good idea how it_s going to perform,_ Fuller confirms. _But these give us a chance, in flight, to modify and fine tune and to get as much bang for our buck as possible._ Specially, they allow engineers to test movements, software and work on any glitches that develop on the real thing.

The first rover is an identical copy of Curiosity; it_s even loaded with the exact same software. The second _ and the one (for obvious reasons) they let me control _ is not fitted with the same onboard computers, earning it the nickname Scarecrow, because it doesn_t have a _brain_. Although it_s the same size as Curiosity, Scarecrow is lighter, to simulate the effects of moving around in the lower gravity of Mars.

Scarecrow is little more than an impressively large remote controlled car. Scroll the tumblers to set distance and angle of movement on the iPhone app, hit go and it reacts instantly, there_s even a stop button if things start going awry.

Thanks to the efforts of its drivers and the engineering team behind them, Curiosity has covered (at the time of writing) more than 300m (1,000ft), including a single push of 42m (138ft), its biggest roll to date since landing on Mars two months ago. Despite being millions of miles away, it_s considerably further than I achieve before grounding Scarecrow.

by Richard Hollingham

__

Top
We use cookies to improve our website. By continuing to use this website, you are giving consent to cookies being used. More details…